Available online at ScienceDirect. Energy Procedia 72 (2015 )

Transcription

1 Available online at ScienceDirect Energy Procedia 72 (2015 ) International Scientific Conference Environmental and Climate Technologies CONECT 2014 How to assess involvement of electricity end user in energy efficiency improvement - analysis of survey results Uldis Bariss*, Aris Dandens, Lelde Timma, Andra Blumberga, Dagnija Blumberga Riga Technical University, Institute of Energy Systems and Environment, Azenes iela 12/1, Riga, LV 1048, Latvia Abstract It have been demonstrated that by providing better information and feedback on electricity consumption to households, it is possible to achieve a considerable level of energy savings. However, factors that influence the decisions of households to engage in energy saving activities often are not clear. This study uses the goal-framing theory to examine the diffusion of energy efficiency measures in society. Based on a survey, logistic regression analysis was performed to determine a combination of factors that would explain the intention to engage in energy efficiency activities. The model which combined normative, gain and hedonic goals, as well as other variables, explained % of deviance and % of adjusted deviance that is lower if compare with similar studies on intention to adopt eco-innovations. Even though there was no clearly dominating goal identified in the goal frame, the normative motivation was slightly more dominant Published The Authors. by Elsevier Published Ltd. by This Elsevier is an open Ltd. access article under the CC BY-NC-ND license ( Peer-review under responsibility of Riga Technical University, Institute of Energy Systems and Environment. Peer-review under responsibility of Riga Technical University, Institute of Energy Systems and Environment Keywords: energy efficiency; goal-framing theory; energy saving behavior; households 1. Introduction Energy efficiency is defined as one of the main targets of the EU to reduce dependency on energy imports, saving on primary energy resources and limiting climate changes. Final household energy consumption is still on the rise and in 2010 accounted for 29.7 % of total electricity consumption in the EU-27 [1]. Therefore efforts to promote energy efficiency in electricity usage are especially important. * Corresponding author. address: uldis.bariss@rtu.lv Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ( Peer-review under responsibility of Riga Technical University, Institute of Energy Systems and Environment doi: /j.egypro

2 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) In electricity supply, it is important to provide consumers with information on their consumption in a timely and informative manner. This feedback will make energy consumption visible, thus establishing knowledge among residential consumers about how much energy they consume and how much they actually pay for energy. Feedback about usage enables consumers to reduce their electricity demand through energy conservation activities that include changing behavior or making energy-efficiency investments in lighting or household appliances. There are a number of pilot projects and research works exploring potential energy savings by providing better information and feedback to households on consumption. For example, Vaasa ETT s report that involved collecting and comparing results from about 100 pilots has shown that immediate feedback resulted in highest energy savings at 8.7 %. The remaining channels for feedback produced almost equal consumption reduction levels. The average saving results by providing detailed invoices were 6 % but providing access to a webpage feedback 5.1 % [2]. The findings from a smart metering study carried out in Latvia have shown that smart meter roll-out and provision of appropriate feedback information can provide potential energy savings at around 11.4 % on average with 8.6 % for a median extrapolating the savings to a general distribution of households [3]. During analysis of factors influencing energy efficiency in this pilot, no correlation was found between energy saving results and such characteristics as general demographic properties, income, education, prior knowledge of energy efficiency measures, interest to be engaged in energy efficiency activities. However, it was identified that respondents who identified before the pilot that it might be possible for them to save energy, did in fact save more during the pilot [4]. This has identified that it is not enough to have knowledge about energy efficiency options but the customers should form an attitude and reach a level of awareness regarding energy efficiency options from their own perspective. This could be analyzed in the framework of innovation diffusion. One of the well-known frameworks of diffusion of innovation has been developed by Rogers [5]. Five stages are identified on how an individual adopts innovation: gains knowledge of an innovation through social networks; forms an attitude towards it; decides to adopt or reject it; implements it; and confirms the decision. In order to promote energy efficiency measures, the generally used practice is to inform the public about possible solutions that can be adopted such as more energy efficient lighting or white appliances. At the same time, in spite of relatively well disseminated information, the implementation of recommended activities are lagging behind. This fact could be explained by the framework of innovation diffusion which states that, after gaining knowledge an individual has to form an attitude towards it. There are a number of theories that are regularly used as a framework to study pro-environmental intentions and behavioral aspects that are relevant during this process. One of these theories is Azjen's [6] theory of planned behavior or on its variations, which was building on Fishbein and Ajzen's theory of reasoned action. The theory defines that the best predictor of actual behavior is the intention to perform such behavior. It is assumed that the intention is to capture the motivation factors that influence actual behavior in the end. The behavioral intention is described as an indication of how hard people are willing to try or how much an effort they are planning to apply in order to adopt the behavior. There are three main determinants of the intention which are defined as attitude towards behavior, subjective norms and perceived control. An attitude towards the adoption of a particular innovation is determined by an individual s beliefs about the consequences of adopting this behavior that will result in a favorable or unfavorable attitude towards the innovation. At the same time, an individual s subjective norms or normative beliefs are determined by his or her perceived expectations of social groups such as family and society and by the person s motivation to conform to these expectations. Perceived control is an individual s view on how easy or difficult it is to perform particular behavior. One of the most important factors in the adoption of pro-environmental innovations is cost. The theory of planned behavior focuses more on an individual s rational choice in the decision-making process. At the same time, pro-environmental activities are often also studied from an altruistic viewpoint. Value-belief-norm theory is one of the frameworks that investigate altruistic intention and behavior and puts more focus on moral

3 272 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) factors for explaining behavior [7]. The theory assumes a hierarchical model where individual value orientations directly influence beliefs, and thereby attitudes and behavior by filtering and evaluating the received information. One s awareness of conceivably harmful consequences and accepted responsibility for these possibly harmful consequences of not behaving pro-environmentally is activating personal norms. They then determine whether a person should engage in a pro-environmental behavior resulting in actions such as environmental policy support, participation in public activities and appropriate personal behaviors [7]. Another framework for pro-environmental behavior is the goal-framing theory developed by Lindenberg and Steg [8]. Goal-framing is a cognitive attitude towards certain situations, which not only affect an individual s choice, but also involves a specific "definition of the situation" - i.e., knowledge of the specific situation of the purposes of the implementation perspective. The framework identified that normative, hedonic, and gain motives are largely determinants of pro-environmental behavior. According to this theory, a normative motive refers to the motivation to act appropriately. Driven by this motive, one is inclined to act according to one s own or others norm s to behave in the right way, and tends to feel morally obligated to display pro-environmental behavior. A hedonic motive is concerned with the motivation to feel better right now such as seeing direct pleasure, seeking personal comfort, or seeking excitement. A gain motive is the goal to guard and improve one s resources [8]. In this theory one goal frame usually is a focal, i.e., actively dominating the individual s perception; the other two are usually in the background. However, the goal frames are interacting between themselves according to particular situations and restraints and the background goals can influence the perception of the central goal frame. This framework is well suited to analyze and develop pro-environmental policy decisions. The most frequently used way to encourage pro-environmental actions is to reduce or even remove the conflict between normative goals, on the one hand, and hedonic and gain goals, for example, by subsidizing environmentally friendly products. However, the study by Steg et al. [9] proposes that strengthening normative goals might be a more sustainable strategy to encourage pro-environmental behavior. This strategy will make individuals focus on the normative consequences of behavioral options, thereby reducing the prevalence of the conflict between hedonic and gain goals, on the one hand, and normative goals, on the other, which can encourage pro-environmental actions. By doing so, people would act pro-environmentally even when hedonic or gain goals are focal that is very important in practice as often there are little benefits on individual s level from pro-environmental actions such as regularly checking your tire pressure, defrosting your freezer or switching off lights when leaving a room. 2. Methodology In order to investigate factors influencing energy efficiency adoption in Latvia, a survey was developed to determine residential customer attitudes to adopt energy efficiency measures. The survey was based on the goalframing theory and there was analyzed all three main elements of goal-framing - the hedonic ("enjoying" the energy efficiency measures), gain (cost savings through energy efficiency measures) and normative (energy efficiency measures to reduce impact on the environment). The overall design of the survey was adopted from Ozaki [10], where the author used it in the research of the behavior of consumers adopting green electricity tariffs. According to Rogers framework of diffusion of innovations at first individuals should gain knowledge of an innovation and at the second stage -- an attitude towards the innovation develops [5]. Therefore, in the work, the survey was distributed via to employees of the utility company Latvenergo which provides energy supply and distribution services assuming that in this population there is well disseminated knowledge about energy efficiency solutions. 3. Data analysis Answers were received from 407 respondents, 387 questionnaires from them was completely filled and was used in further data analysis. Since the survey was distributed among utility employees the sample group was biased by male respondents, people with higher education and average income per household member, see Table 1. These

4 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) biases could affect the representativeness of the survey. However, the study s intention was to explore the attitudes only of those people, who have good knowledge about energy efficiency and are in the network of such people in order to analyze how in these circumstances attitude towards energy efficiency are formed. Additionally, pilot studies have not identified correlation between gender and education level from one side and engagement into energy saving behavior from the other [4]. Table 1. Demographic of population surveyed and comparison to average in Latvia. Parameter, unit Statistics from questionnaire Average in Latvia, (year) Age, years on average (2011) Gender, % males (2014) Income per household member, EUR/month (2012) Household size, persons per household (2013) Education % higher (2011) % secondary (2011) Housing type, % living in flat (2013) Language spoken at home, % speaks Latvian (2011) Correlation and logistic regression analysis approached was adopted from Vigants et al. [11] to perform analysis of answers to questionnaire. Correlation analysis was performed to explore how goal frames might affect energy efficiency behavior. The answers from the questionnaire where combined under the following groups: normative, hedonic, gain and mixed goals. Correlation analysis was carried out between the answer about whether the respondent is actively engaged in energy saving activities and all other answers of the questionnaire (independent variable). The correlation coefficient (R) and mean value was calculated for each independent variable at 95 % confidence level for both groups who are and who are not involved in energy saving, thus allowing to compare statistically different answers across these two groups. The interpretation of the correlation coefficient was used in the same manner as in the study on the adoption intention of green electricity tariff by Ozaki [10] with the reference to Cohen [12]; where R in the boundaries ± is referred as small correlations for behavioral studies, ± medium, ± large correlations. Logistic regression analysis was performed to determine a combination of factors that explain the intention to use or not to use energy efficiency measures. Only the statistically significant at the confidence level of 95 % combinations of answers from each of normative, hedonic, gain and mixed goals groups were then combined into one logistic regression model. For logistic regression analysis, a model was created, where dependent variables describe events with two possible statistics the respondent is engaged in energy saving activities or not. For logistic regression analyses respondents who stated that they are using energy efficiency measures but are not actively involved in this activity where also included as it gave better predictability for the model. Possible outcomes are modelled as a function of the predictor variables, using a logistic function, expressed by Equation (1):, (1) where P outcome probabilities if independent variable is X, as independent variables X are used answers from the survey; _0 constant of the fitted model; _1, _2 _k estimated coefficients.

5 274 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) Stepwise backward selection factor selection was chosen with P-value to enter To estimate the accuracy of the fitted model, the percentage of deviance explained by the model (R 2 ) was calculated according to Equation (2)., (2) Adjusted deviance (R 2 adj) was calculated in similar manner to R-squared, see Equation (3)., (3) where p equals the number of coefficients in the fitted model, including constant term. 4. Results Correlation analysis shows that the normative view was expressed more frequently in both groups respondents who are actively applying energy efficiency measures and respondents who are not actively involved or do not take energy efficiency measures at all, see Table 2. Table 2. Mean scores and correlation coefficients between goal frames and energy efficiency behavior. Question Goal* Engaged in energy efficiency Not engaged in energy efficiency R** Mean*** R** Mean*** When I use energy efficiency appliances I'm positively contributing to the environment N (6) (6) Usage of energy efficient appliances reduce pollution of the environment N (6) (6) There will be an ecological catastrophe if we will continue to act as we have done up until now N (6) (6) Environmental pollution might become a serious issue for you and your family N (6) (6) My closest people are also using energy efficiency appliances H (6) (6) Energy efficient appliances are easy to acquire and use H (6) (6) Electricity prices should be as low as possible G (6) (6) Energy efficient appliances are more expensive G (6) (6) * N normative goal, H hedonic goal, G gain goal. ** Answers with correlation coefficients in a range ± (medium) or ± (large) have been included provided they are statistically significant at 95 % confidence level and this condition is satisfied at least in a one of subgroups. *** Values in the scale (6) as: 1 strongly disagree, 3 disagree, 4 agree, 6 strongly agree. The main similarities between the two groups include: awareness of beneficial effect of energy efficient electrical appliances to the environment and acknowledgement that environmental pollution might become a serious issue not only for general public but also for each individual. However, respondents who are engaged in energy saving have expressed slightly stronger pro-environmental values and norms. For example, their answer on the statement When I use energy efficiency appliances I'm positively contributing to environment has an average mean 5.19 on the scale of 6 when the group not engaged in energy savings has an average mean It has been demonstrated by Steg at al. [9] that normative values and goals, such as the need to prevent global warming and water pollution, willingness to accept restrictions in every-day life, support for large scale business and political restrictions clearly correlates with the attitude for implementation of energy efficiency measures.

6 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) The tendency identified in this survey is that, although normative beliefs are more dominant in both groups, gain and hedonic intentions are also represented as background motives. At the same time, this tendency is slightly more observable in the group of respondents who are engaged in energy savings. The results for the logistic regression models fitted for individual motivations shows that there is no one clearly dominating motivation. The final model Energy efficiency motivation (which combines normative, gain, hedonic motivation and other variables together) explained a % deviance and % of adjusted deviance; see Table 3. Table 3. Analysis of deviance and residuals for the logistic regression model Energy efficiency motivation. The model Explained by the model the sum of squared deviations, % Explained by the model the sum of squared deviations, adjusted, % (p a ) Normative motivation **** (27) (6) Gain motivation ** (18) (3) Hedonic motivation **** (16) (3) Other variables **** (16) (3) Energy efficiency motivation **** (37) (9) **Significant at P < 0.01; ***Significant at P < 0.001; ****Significant at P < a p = the number of coefficients in the fitted model, including the constant term. b f = the number of factors in the fitted model. Mean square error, 10-3 (f b ) The statistical model of Energy efficiency motivation (EE USE ) is given as Eq. (4). EE USE exp( /(1 exp( ) (4) Where the is calculated as given in Eq. (5) X ( X 2( X 3( X 4( X 5( X 6( X 7( X 8( 1 (5) Where the acronyms and values for estimates used in the Eq. (5) are given in Table 4. Table 4. Logistic regression parameters and estimates for the model Energy efficiency motivation. Parameter* Estimate Parameter* Estimate Parameter* Estimate Parameter* Estimate if X1 = if X3 = if X5= if X7= if X1 = if X3 = if X5= if X7= if X1 = if X3 = if X5= if X7= if X1 = if X3 = if X5= if X7= if X1 = if X3 = if X5 = if X7= if X2 = if X4 = if X6= if X8= if X2 = if X4= if X6= if X2 = if X4= if X6= if X2 = if X4= if X6= if X2 = if X4= if X6= *Respondents answers to the following statements or questions as follows: X1 energy-efficient appliances for domestic use is less effective than the traditional home electrical equipment;

7 276 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) X2 nature will be able to retain a balance in order to cope with industrial nations impact on environment; X3 electricity costs should be as low as possible; X4 when I use energy-efficient electrical equipment, I am doing something good for the environment; X5 local environmental pollution has to be reduced; X6 environmental pollution has to be reduced; X7 I have a personal responsibility to do everything to prevent environmental pollution; X8 whether concern for the environment is for you the main motivator to use energy-efficient appliances. Values in the scale of six were given for each question as 1 strongly disagree, 3 disagree, 4 agree, 6 strongly agree. For the question X8 in the scale 0 no, I don t, 1 yes, I do. Based on the answers in the logistic regression model (Eq. (5)) is obtained. For example, in the case when the respondent gives answer that he strongly disagrees (scale 1) with the question X2, then the estimate for is ; see Table 4. If we were to compare results given in Table 3 with similar studies, then the adjusted deviance in our case study is lower, since Vigants et al. [11] explained % of adjusted deviance when the intention to adopt eco-innovations was studied, and Tonglet et al. [13] explained 33.3 % of adjusted deviance when the intention to recycle was studied. This can be due to the fact that the context of the survey was embracing all energy efficiency activities applicable to households and this scope might be too broad. Goals can change not only due to underlying values but also due to situational cues that imply that reactions are very context specific [9]. As argued by Lindenberg and Stag [14] goals can become focal as an automatic reaction to cues and when some of them become focal, they change our behavior affecting what we focus on, what information we are sensitive to, what information we neglect, what knowledge and what concepts are being activated at a given moment, what we like and dislike, what we expect others to do. Consequently respondents might not be placed in a specific enough context in order for all of them to form a more definite attitude towards expected behavior that resulted in more scattered results among different goals. 5. Conclusion The results for the logistic regression models fitted for individual motivations shows that there is no one very clearly dominating motivation; however, normative motivation was observed only slightly more frequently. It was clearly dominating in correlation analysis in both groups respondents who are and who are not actively involved in energy efficiency activities. The final model Energy efficiency motivation, which combines normative, gain, hedonic motivation and other variables together, explained % of deviance and % of adjusted deviance that is lower if compared with similar studies on intention to adopt eco-innovations. This might be due to the broad survey setting; therefore, the predictability of this methodical approach could be improved by narrowing the survey to a very specific energy efficiency activity where respondents are engaged such as purchasing only LED or A+++ white appliance. Acknowledgements The work has been supported by the National Research Program Energy efficient and low-carbon solutions for a secure, sustainable and climate variability reducing energy supply (LATENERGI). References [1] EC Joint Research centre, Energy Efficiency Status Report 2012 // [ ] [2] VaasaEET (2011). The Potential of Smart Meter Enabled Programs to Increase Energy and Systems Efficiency: A Mass Pilot Comparison Empower Demand. Report for European Smart Metering Industry Group.

8 Uldis Bariss et al. / Energy Procedia 72 ( 2015 ) [3] Bariss, U., Kamenders, A., Vitolins, V., Blumberga, D. Energy efficiency results of Smart metering pilot in a context of Cost Benefit analysis of smart meters in Latvia. ECOS 2014; 27st International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems; Conference Proceedings, Turku, Finland, June 15 19, 2014, p [4] Bariss, U., Kuznecova, T., Laicane, I., Blumberga, D. Analysis of Factors Influencing Energy Efficiency in a Smart Metering Pilot. Energetika, 2014; 60: [5] Rogers EM. Diffusion of Innovations, 5th edn. Free Press: New York; [6] Aijzen., I. The theory of planned behavior. Original Organizational Behavior and Human Decision Processes, 1991; 50: [7] Stern, P. C., Dietz, T., Abel, T., Guagnano, G. A., & Kalof, L.. A value-belief-norm theory of support for social movements: the case of environmentalism. Research in Human Ecology 1999; 6: [8] Lindenberg S., Steg L. Normative, Gain and Hedonic Goal Frames Guiding Environmental Behavior, Journal of Social 2007; 63(1): [9] Steg L., Bolderdijk J.W., Keizer K., Perlaviciute G. An Integrated Framework for Encouraging Pro-environmental Behaviour: The role of values, situational factors and goals, Journal of Environmental Psychology 2014; 38: [10] Ozaki R. Adopting sustainable innovation: what makes consumers sign up to green electricity? Business Strategy and the Environment 2011; 20: [11] V gants E., Blumberga A., jabs I., Timma L. The dynamics of technological substitution: the case of eco-innovations diffusion, Journal of Cleaner Production, Under peer-review, [12] Cohen J.,. Statistical Power Analysis for the Behavioural Science, 2nd Edition. Erlbaum: Hillsdale, NJ [13] Tonglet M., Phillips P.S., Bates M. Determining the drivers for householderpro-environmental behaviour: waste minimization compared to recycling, Resources, Conservation and Recycling 2004; 42: [14] Lindenberg, S. and Steg, L. Goal-framing Theory and Norm-Guided Environmental Behavior. In H. van Trijp (ed.), Encouraging Sustainable Behavior, New York: Psychology Press p